Prevention of Endophthalmitis after Cataract Surgery: Making the Most of the Evidence

Abstract

Concern about preventing endophthalmitis after cataract surgery pervades and controls many of our perioperative behaviors. Such behaviors include multiple different approaches to preoperative use of topical antibiotics and antiseptics, intraoperative and/or postoperative delivery of antibiotics, and type and location of wound construction and closure. There is tremendous variability in such behaviors, although many are practiced with fundamentalist fervor. Such variation exists across individuals, centers, regions, and countries. In addition, endophthalmitis after cataract surgery is relatively rare, thus making it almost impossible for an individual surgeon or even a practice to extrapolate from its own local clinical experience. Methodologically sound clinical research on rates, risk factors, and prevention of endophthalmitis is also very difficult to perform, not only because of the rarity of the condition, but also because of the multiple potential variables that may be important to record. How, then, should one approach the evidence available to date and incorporate it into practice?Evidence-based practice, as David Sackett succinctly stated, is “about integrating individual clinical expertise and the best external evidence.”1Sackett D.L. Rosenberg W.M. Gray J.A. et al.Evidence based medicine: what it is and what it isn’t.BMJ. 1996; 312: 71-72Crossref PubMed Google Scholar Of course, we will never have large randomized clinical trials to address every clinical question or scenario of interest. Yet, one can create a hierarchy of methodologic approaches to patient-oriented research on endophthalmitis and see what inferences can then be brought to bear on clinical practice. I make no pretense to include all the methodologies and research published, but will take the liberty of presenting my own interpretation.First, it is generally acknowledged that prospective randomized clinical trials offer the highest level of evidence. In this domain, I believe that there is good evidence that the use of povidone–iodine in the preparation of the ocular surface is superior to the use of soap and irrigation.2Speaker M.G. Menikoff J.A. Prophylaxis of endophthalmitis with topical povidone-iodine.Ophthalmology. 1991; 98: 1769-1775Abstract Full Text PDF PubMed Scopus (413) Google Scholar It is unknown, however, whether povidone–iodine is superior to other antiseptics, such as the 0.05% chlorhexidene routinely used in the study from Sweden reported in this issue.3Lundström M. Wejde G. Stenevi U. et al.Endophthalmitis after cataract surgery A nationwide prospective study evaluating incidence in relation to incision type and location.Ophthalmology. 2007; 114: 866-870Abstract Full Text Full Text PDF PubMed Scopus (251) Google Scholar A recent publication4Barry P. Seal D.V. Gettinby G. et al.ESCRS study of prophylaxis of postoperative endophthalmitis after cataract surgery: preliminary report of principal results from a European multicenter study.J Cataract Refract Surg. 2006; 32: 407-410Abstract Full Text Full Text PDF PubMed Scopus (313) Google Scholar has reported preliminary findings from a randomized clinical trial on endophthalmitis prevention from a multicenter study in Europe. This study provides strong evidence that the application of 1 mg of cefazolin intracamerally at the end of surgery reduced the rate of endophthalmitis approximately 5-fold and called into question the use of perioperative topical antibiotics. Does this mean that we should all immediately adopt the use of intracameral cefazolin? After all, the finding of a 5-fold risk reduction in a randomized clinical trial is very strong. Unfortunately, the inference is not so straightforward. The rate of endophthalmitis in that study’s control group was higher than 1 in 300 cases—extraordinarily high. The rate of endophthalmitis in the treated group was approximately 1 in 1400. If your/your center’s rate is currently at 1 in 1500 or less, does that mean that you might expect a further significant reduction if you used intracameral cefazolin? Or, does this mean that intracameral cefazolin is particularly helpful in situations where other methods of preventing infection are not used or are inadequate? This is an open question that is worth answering. There has also been a randomized clinical trial, although from a single center, that has examined endophthalmitis rates by intraocular lens (IOL) material and incision site.5Nagaki Y. Hayasaka S. Kadoi C. et al.Bacterial endophthalmitis after small-incision cataract surgery: effect of incision placement and intraocular lens type.J Cataract Refract Surg. 2003; 29: 20-26Abstract Full Text Full Text PDF PubMed Scopus (202) Google Scholar This study showed no difference in disease rates by IOL material for either corneal or scleral incisions, and about a 3-fold excess risk for corneal versus scleral incision sites. Inference from that study is straightforward, at least as it applies to the specific surgical techniques used.I regard data from large representative samples as secondmost reliable in the hierarchy of evidence regarding endophthalmitis. The Lundström et al article in this issue3Lundström M. Wejde G. Stenevi U. et al.Endophthalmitis after cataract surgery A nationwide prospective study evaluating incidence in relation to incision type and location.Ophthalmology. 2007; 114: 866-870Abstract Full Text Full Text PDF PubMed Scopus (251) Google Scholar is an excellent example of the strengths of this kind of research. The data are national, unbiased, and collected in a prospective manner, with very clear definitions of the outcome and potential risk factors. This research has some very definitive findings that I find compelling and instructive, particularly because Sweden has both a lower rate of endophthalmitis and some important practices that differ from those routinely employed in the United States. The most notable distinction is the lack of routine use of preoperative or postoperative topical antibiotics in Sweden. In the U.S., almost everyone uses topical antibiotics both before surgery (starting either several days in advance or in the minutes just before surgery) and for about a week after. However, this common practice has no sound evidence base. One can try only to extrapolate from animal studies or small series that measure antibiotic levels in the anterior chamber (AC) at various intervals. This sort of indirect evidence pales in comparison to that offered by the European randomized trial4Barry P. Seal D.V. Gettinby G. et al.ESCRS study of prophylaxis of postoperative endophthalmitis after cataract surgery: preliminary report of principal results from a European multicenter study.J Cataract Refract Surg. 2006; 32: 407-410Abstract Full Text Full Text PDF PubMed Scopus (313) Google Scholar and the national Swedish study, which both suggest that perioperative topical antibiotics confer minimal or no benefit, at least when intracameral cefazolin is used. Cefazolin, long off patent, is very inexpensive. I infer that the perioperative use of topical antibiotics is of questionable effectiveness in reducing endophthalmitis rates, is doubtless cost-ineffective, and is certainly a practical burden for many elderly patients.Third in the methodologic hierarchy is the case–control study, the most practical way to study risk factors for relatively rare events. Fortunately, the results of such studies have been quite consistent with the large population databases. Specifically, consistent findings have been excess risk associated with corneal incisions, age (especially over 80 years), and loss of posterior capsular integrity (i.e., direct access of bacteria to the vitreous). When such findings are consistent across study design and populations, I find them very convincing. Of course, some risk factors are modifiable, and others not. The population of patients over 80 undergoing cataract surgery will increase dramatically over the next few decades, based on our national census projections. In the absence of improved prophylaxis, this demographic trend alone would be expected to increase endophthalmitis rates in the U.S.6West E.S. Behrens A. McDonnell P.J. et al.The incidence of endophthalmitis after cataract surgery among the U.S. Medicare population increased between 1994 and 2001.Ophthalmology. 2005; 112: 1388-1394Abstract Full Text Full Text PDF PubMed Scopus (242) Google Scholar Capsular rupture is not inherently modifiable, given that all surgeons try already to avoid it. However, it is such an important and consistent risk factor that its presence indicates that one might consider modifying routine care in this setting. This would be especially applicable if the capsular rupture were to occur in a patient over 85, a situation in which two strong independent risk factors coexist. One way to do so would be to increase the intensity of prophylaxis. For example, one might use intracameral cefazolin in this context, if that were not already the routine. Additionally, one might take extra precautions with the wound in such a patient (e.g., by adding a stitch to a corneal wound). Or, one might choose to examine such a patient more frequently over the early postoperative period to detect and treat potential endophthalmitis more effectively.There are two opportunities for bacteria to gain access to the inside of the eye, either at the time of the surgery or in the early postoperative period, before epithelialization of the wound. It is not logical that the first mechanism should be more or less associated with wound type or location. Presumably, the excess risk for corneal incisions relates to influx of ocular surface bacteria within the first postoperative day. Does the consistent finding of excess risk with corneal incisions mean that this is an unsafe procedure? There are many surgeons who have performed thousands of consecutive cataract surgeries using corneal incisions without a single case of endophthalmitis. My own interpretation of the finding, merged with clinical experience, is that, on average, corneal incisions increase risk to a modest degree. However, that excess risk can be mitigated by expertise (i.e., close attention to wound construction and integrity) or by other factors that might reduce endophthalmitis risk, such as brief operative time or other changes in procedure.Given my simplistic view of endophthalmitis pathophysiology—ocular surface bacteria gaining entrance to the eye either during surgery or within the first day or so postoperatively—it is straightforward to create a wish list for an antibiotic prophylaxis strategy. First, the antibiotic should cover the well-known commonly acquired bacteria. Second, the intraocular concentration (ideally, in both the AC and vitreous) of that antibiotic should kill those same bacteria, both at the close of surgery and for a duration of 24 to 48 hours. Third, the antibiotic strategy must have an extremely low adverse event rate. This is obviously crucial, because one would envision exposing a large population to the strategy on a routine basis, and the complication to be prevented is relatively rare. For these same reasons, the intervention must be cost-effective. Fourth and finally, the intervention must be studied in well-controlled patient populations, so that the evidence base for its use is strong enough to merit widespread adoption. Topical antibiotics do not meet such criteria. The use of intracameral cefazolin meets some of these criteria, but its duration of activity in the eye is too brief. Perhaps the strategy that comes the closest to meeting the criteria that are currently available is the perioperative use of oral moxifloxacin, a drug with excellent intraocular penetration, bacterial coverage, and ocular pharmacokinetics with just a few perioperative doses.7Hariprasad S.M. Shah G.K. Mieler W.F. et al.Vitreous and aqueous penetration of orally administered moxifloxacin in humans.Arch Ophthalmol. 2006; 124: 178-182Crossref PubMed Scopus (106) Google Scholar However, this approach has not been evaluated for endopthalmitis prevention in a large patient population. What about the future? One can readily imagine the placement of an antibiotic depot directly into the eye at the end of surgery. Such a depot would release an effective and safe antibiotic over the initial hours and days after surgery and, presumably, lower endophthalmitis rates in a safe and predictable manner. Of course, this strategy, though appealing and logical, would also require validation in a large patient population. Such trials are feasible, despite the large sample sizes required, because cataract surgery is so common and the outcome of interest, endophthalmitis, is straightforward to identify and count. Concern about preventing endophthalmitis after cataract surgery pervades and controls many of our perioperative behaviors. Such behaviors include multiple different approaches to preoperative use of topical antibiotics and antiseptics, intraoperative and/or postoperative delivery of antibiotics, and type and location of wound construction and closure. There is tremendous variability in such behaviors, although many are practiced with fundamentalist fervor. Such variation exists across individuals, centers, regions, and countries. In addition, endophthalmitis after cataract surgery is relatively rare, thus making it almost impossible for an individual surgeon or even a practice to extrapolate from its own local clinical experience. Methodologically sound clinical research on rates, risk factors, and prevention of endophthalmitis is also very difficult to perform, not only because of the rarity of the condition, but also because of the multiple potential variables that may be important to record. How, then, should one approach the evidence available to date and incorporate it into practice? Evidence-based practice, as David Sackett succinctly stated, is “about integrating individual clinical expertise and the best external evidence.”1Sackett D.L. Rosenberg W.M. Gray J.A. et al.Evidence based medicine: what it is and what it isn’t.BMJ. 1996; 312: 71-72Crossref PubMed Google Scholar Of course, we will never have large randomized clinical trials to address every clinical question or scenario of interest. Yet, one can create a hierarchy of methodologic approaches to patient-oriented research on endophthalmitis and see what inferences can then be brought to bear on clinical practice. I make no pretense to include all the methodologies and research published, but will take the liberty of presenting my own interpretation. First, it is generally acknowledged that prospective randomized clinical trials offer the highest level of evidence. In this domain, I believe that there is good evidence that the use of povidone–iodine in the preparation of the ocular surface is superior to the use of soap and irrigation.2Speaker M.G. Menikoff J.A. Prophylaxis of endophthalmitis with topical povidone-iodine.Ophthalmology. 1991; 98: 1769-1775Abstract Full Text PDF PubMed Scopus (413) Google Scholar It is unknown, however, whether povidone–iodine is superior to other antiseptics, such as the 0.05% chlorhexidene routinely used in the study from Sweden reported in this issue.3Lundström M. Wejde G. Stenevi U. et al.Endophthalmitis after cataract surgery A nationwide prospective study evaluating incidence in relation to incision type and location.Ophthalmology. 2007; 114: 866-870Abstract Full Text Full Text PDF PubMed Scopus (251) Google Scholar A recent publication4Barry P. Seal D.V. Gettinby G. et al.ESCRS study of prophylaxis of postoperative endophthalmitis after cataract surgery: preliminary report of principal results from a European multicenter study.J Cataract Refract Surg. 2006; 32: 407-410Abstract Full Text Full Text PDF PubMed Scopus (313) Google Scholar has reported preliminary findings from a randomized clinical trial on endophthalmitis prevention from a multicenter study in Europe. This study provides strong evidence that the application of 1 mg of cefazolin intracamerally at the end of surgery reduced the rate of endophthalmitis approximately 5-fold and called into question the use of perioperative topical antibiotics. Does this mean that we should all immediately adopt the use of intracameral cefazolin? After all, the finding of a 5-fold risk reduction in a randomized clinical trial is very strong. Unfortunately, the inference is not so straightforward. The rate of endophthalmitis in that study’s control group was higher than 1 in 300 cases—extraordinarily high. The rate of endophthalmitis in the treated group was approximately 1 in 1400. If your/your center’s rate is currently at 1 in 1500 or less, does that mean that you might expect a further significant reduction if you used intracameral cefazolin? Or, does this mean that intracameral cefazolin is particularly helpful in situations where other methods of preventing infection are not used or are inadequate? This is an open question that is worth answering. There has also been a randomized clinical trial, although from a single center, that has examined endophthalmitis rates by intraocular lens (IOL) material and incision site.5Nagaki Y. Hayasaka S. Kadoi C. et al.Bacterial endophthalmitis after small-incision cataract surgery: effect of incision placement and intraocular lens type.J Cataract Refract Surg. 2003; 29: 20-26Abstract Full Text Full Text PDF PubMed Scopus (202) Google Scholar This study showed no difference in disease rates by IOL material for either corneal or scleral incisions, and about a 3-fold excess risk for corneal versus scleral incision sites. Inference from that study is straightforward, at least as it applies to the specific surgical techniques used. I regard data from large representative samples as secondmost reliable in the hierarchy of evidence regarding endophthalmitis. The Lundström et al article in this issue3Lundström M. Wejde G. Stenevi U. et al.Endophthalmitis after cataract surgery A nationwide prospective study evaluating incidence in relation to incision type and location.Ophthalmology. 2007; 114: 866-870Abstract Full Text Full Text PDF PubMed Scopus (251) Google Scholar is an excellent example of the strengths of this kind of research. The data are national, unbiased, and collected in a prospective manner, with very clear definitions of the outcome and potential risk factors. This research has some very definitive findings that I find compelling and instructive, particularly because Sweden has both a lower rate of endophthalmitis and some important practices that differ from those routinely employed in the United States. The most notable distinction is the lack of routine use of preoperative or postoperative topical antibiotics in Sweden. In the U.S., almost everyone uses topical antibiotics both before surgery (starting either several days in advance or in the minutes just before surgery) and for about a week after. However, this common practice has no sound evidence base. One can try only to extrapolate from animal studies or small series that measure antibiotic levels in the anterior chamber (AC) at various intervals. This sort of indirect evidence pales in comparison to that offered by the European randomized trial4Barry P. Seal D.V. Gettinby G. et al.ESCRS study of prophylaxis of postoperative endophthalmitis after cataract surgery: preliminary report of principal results from a European multicenter study.J Cataract Refract Surg. 2006; 32: 407-410Abstract Full Text Full Text PDF PubMed Scopus (313) Google Scholar and the national Swedish study, which both suggest that perioperative topical antibiotics confer minimal or no benefit, at least when intracameral cefazolin is used. Cefazolin, long off patent, is very inexpensive. I infer that the perioperative use of topical antibiotics is of questionable effectiveness in reducing endophthalmitis rates, is doubtless cost-ineffective, and is certainly a practical burden for many elderly patients. Third in the methodologic hierarchy is the case–control study, the most practical way to study risk factors for relatively rare events. Fortunately, the results of such studies have been quite consistent with the large population databases. Specifically, consistent findings have been excess risk associated with corneal incisions, age (especially over 80 years), and loss of posterior capsular integrity (i.e., direct access of bacteria to the vitreous). When such findings are consistent across study design and populations, I find them very convincing. Of course, some risk factors are modifiable, and others not. The population of patients over 80 undergoing cataract surgery will increase dramatically over the next few decades, based on our national census projections. In the absence of improved prophylaxis, this demographic trend alone would be expected to increase endophthalmitis rates in the U.S.6West E.S. Behrens A. McDonnell P.J. et al.The incidence of endophthalmitis after cataract surgery among the U.S. Medicare population increased between 1994 and 2001.Ophthalmology. 2005; 112: 1388-1394Abstract Full Text Full Text PDF PubMed Scopus (242) Google Scholar Capsular rupture is not inherently modifiable, given that all surgeons try already to avoid it. However, it is such an important and consistent risk factor that its presence indicates that one might consider modifying routine care in this setting. This would be especially applicable if the capsular rupture were to occur in a patient over 85, a situation in which two strong independent risk factors coexist. One way to do so would be to increase the intensity of prophylaxis. For example, one might use intracameral cefazolin in this context, if that were not already the routine. Additionally, one might take extra precautions with the wound in such a patient (e.g., by adding a stitch to a corneal wound). Or, one might choose to examine such a patient more frequently over the early postoperative period to detect and treat potential endophthalmitis more effectively. There are two opportunities for bacteria to gain access to the inside of the eye, either at the time of the surgery or in the early postoperative period, before epithelialization of the wound. It is not logical that the first mechanism should be more or less associated with wound type or location. Presumably, the excess risk for corneal incisions relates to influx of ocular surface bacteria within the first postoperative day. Does the consistent finding of excess risk with corneal incisions mean that this is an unsafe procedure? There are many surgeons who have performed thousands of consecutive cataract surgeries using corneal incisions without a single case of endophthalmitis. My own interpretation of the finding, merged with clinical experience, is that, on average, corneal incisions increase risk to a modest degree. However, that excess risk can be mitigated by expertise (i.e., close attention to wound construction and integrity) or by other factors that might reduce endophthalmitis risk, such as brief operative time or other changes in procedure. Given my simplistic view of endophthalmitis pathophysiology—ocular surface bacteria gaining entrance to the eye either during surgery or within the first day or so postoperatively—it is straightforward to create a wish list for an antibiotic prophylaxis strategy. First, the antibiotic should cover the well-known commonly acquired bacteria. Second, the intraocular concentration (ideally, in both the AC and vitreous) of that antibiotic should kill those same bacteria, both at the close of surgery and for a duration of 24 to 48 hours. Third, the antibiotic strategy must have an extremely low adverse event rate. This is obviously crucial, because one would envision exposing a large population to the strategy on a routine basis, and the complication to be prevented is relatively rare. For these same reasons, the intervention must be cost-effective. Fourth and finally, the intervention must be studied in well-controlled patient populations, so that the evidence base for its use is strong enough to merit widespread adoption. Topical antibiotics do not meet such criteria. The use of intracameral cefazolin meets some of these criteria, but its duration of activity in the eye is too brief. Perhaps the strategy that comes the closest to meeting the criteria that are currently available is the perioperative use of oral moxifloxacin, a drug with excellent intraocular penetration, bacterial coverage, and ocular pharmacokinetics with just a few perioperative doses.7Hariprasad S.M. Shah G.K. Mieler W.F. et al.Vitreous and aqueous penetration of orally administered moxifloxacin in humans.Arch Ophthalmol. 2006; 124: 178-182Crossref PubMed Scopus (106) Google Scholar However, this approach has not been evaluated for endopthalmitis prevention in a large patient population. What about the future? One can readily imagine the placement of an antibiotic depot directly into the eye at the end of surgery. Such a depot would release an effective and safe antibiotic over the initial hours and days after surgery and, presumably, lower endophthalmitis rates in a safe and predictable manner. Of course, this strategy, though appealing and logical, would also require validation in a large patient population. Such trials are feasible, despite the large sample sizes required, because cataract surgery is so common and the outcome of interest, endophthalmitis, is straightforward to identify and count. ErratumOphthalmologyVol. 114Issue 6PreviewWith apologies from the author, editors, and reviewers, in the May editorial entitled “Prevention of Endophthalmitis after Cataract Surgery: Making the Most of the Evidence” (2007;114:831–2) cefuroxime should replace cefazolin every time it is mentioned. Full-Text PDF